Educational Performance and Cultural Consumption: Some Evidence from Oecd Teenagers

EDUCATIONAL PERFORMANCE AND CULTURAL CONSUMPTION: SOME EVIDENCE FROM OECD TEENAGERS

V. Fernández-Blanco* M. García-Diez* J. Prieto-Rodríguez*

*Departamento de Economía - Universidad de Oviedo - Spain

ABSTRACT

The aim of this paper is to analyse the young people’s cultural consumption habits in the OECD countries. There are many studies that describe the characteristics of cultural goods consumers in different countries (i. e. Baumol and Bowen 1966; Throsby and Withers, 1979; Towse, 1994, O’Hagan, 1996) but they are specially related to adult people. On the other hand, there are many opinions arguing that young people’s cultural consumption could be very important since in some sectors, such as movies, they represent a relevant market share and, in any case, they are the audience of the future. However, the researches devoted to this group of population are very scarce.

Our paper is focused on this group of population, specifically in fifteen years old teenagers. Using a database coming from the PISA (Programme for Indicators of Student Achievement) Project, that in 2000 includes a survey with a sample of around 250.000 teenagers coming from 27 OECD and 4 non-OECD countries, we make a comparative analysis of their current consumption of some cultural commodities (movies, museums and art galleries, theatre and other performing arts, assistance to popular music concerts and to sport events). This database also gives us relevant information on the socio-economic characteristics of the teenagers and their families (sex, level of studies, size of the family,…), their educational performing and their parents’ compromise with their cultural and educative activities. Using all this information, we estimate some multiple discrete choice econometric models to characterise the young audience of each kind of cultural good in each country, and to discover if there are relevant differences or common features among them.

1 INTRODUCTION

Tastes and taste formation are very common discussing topics in Cultural Economics. In general, there are two approaches to tastes on this field. First, originally inspired by Stigler and Becker (1977), considers tastes as a constant and, more or less, all people have the same preferences. Hence we cannot find changes in tastes but changes in their form of expression and in people’s consuming skills (McCain, 1981, 1995 y 2003). It is important to notice that, following this approach, paternalism or any other policy devoted to guide individual consumption are useless. Second, tastes are the outcome of a dynamic process of learning by consuming (Lévy-Garboua and Montmarquette, 1996). In this case, cultural goods are experience goods and their taste is able to be changed through repeated exposure and experience (Lévy-Garboua, 2003) and, of course, policy decisions could be very important in the cultivation of taste process.

Childhood exposure to the arts is close to these ideas. Obviously, children are the future demand of any cultural good and they are in a crucial moment of their taste formation process, where participation and exposure to the arts and education in general would play the main roles. However there are only a few studies on children exposure to the arts (Morrison and West, 1986; Abbé-Decarroux, 1995) and their outcomes are not very conclusive: for instance, Morrison and West (1986, p. 157) point out that “child participation is a powerful predictor of adult theatre attendance” but they cannot reach a similar conclusion on exposure. Under these circumstances, McCain (2003, p.446) says that “these studies may be an indirect evidence that cultivation of taste is a real possibility”.

Our aim is to achieve some new empirical evidence about young people’s consumption of certain kinds of cultural goods1. We will pay special attention, on the one hand, to the effect of education and, in particular, to the students’ performance in the education system; and, on the other hand, to the influence of familiar environment on young people’s consumption choices. We suppose that both factors affect the taste formation process and, even, that the relative’s consumption generates some external effects that could influence teenagers’ behaviour. As usual, we have also included a set of socio-economic variables to control personal characteristics of the individuals or their family (as gender, marital status, familiar responsibilities, relationship with the economic activity, and so on).

1 Studies that analyses the characteristics of adult cultural consumers are more frequent (i. e. Baumol and Bowen 1966; Throsby and Withers, 1979; Towse, 1994, O’Hagan, 1996) 2 Moreover, we pretend to include in our paper some international comparisons using the same cohort of people (fifteen years old teenagers) to analyse the situation in a group of countries that can represent different cultures, geographical areas and levels of economic development. With this procedure, we want to achieve other interesting target: testing if familiar environment and educational level influences are more or less the same among this group of countries and even if they are independent from other features as, for instance, the degree of economic development or the family income.

DATA

To achieve our aims we use date coming from the Survey of the OECD’s Programme for International Student Assessment (PISA). It declared objective is “to measure how young adults, at age 15 and therefore the end of compulsory schooling, are prepared to meet the challenges of today’s knowledge societies” (OECD, 2001, p 14); specifically, it wanted to know how students can use what they have learned in reading, mathematics and science. In 2000 a PISA survey was conducted to a sample of more than a quarter of a million students, representing almost 17 million 15-years-olds enrolled in the schools of the 32 participating countries (the 28 OECD countries plus Brazil, Latvia, Liechtenstein and Russia). The majority of the questions of this survey were related to educational items, but there were also some questions about the socio-economic characteristics of the student and his/her family, such as gender, economic and social background, and his/her activities at home and school, including some leisure and cultural activities like attending to pictures, performing arts, pop concerts or sporting events. So, although PISA is a programme oriented to educational aspects, it also offers some information about cultural performance of 15-years-olds and the possible influence of their parents and relatives and their familiar environment. Specifically, it gives us data on young people’s attendance to the cinemas, to the theatres or to the concerts of music, and it also collects information about other cultural or leisure habits (as reading or going to live sports). Related to the familiar environment, a variable that it is specially relevant in our research, this Survey offers us information about the parents’ educational level and other cultural behaviours, specially if the interviewee consumes cultural products jointly with her parents or if the family talk about cultural matters and devices.

Due to we have a lot of information, we have selected only two cultural goods and seven countries. First, we have chosen attendance to cinema and to theatre. Both commodities are consumed outside home, usually they have been presented as possible substitutive goods and, finally, the former can be considered as a market-oriented industrial good, while the last is more linked to the non-profit sector. Second, we have chosen nine countries that we consider as 3 representatives of different cultures and levels of economic development. Then, we have included USA; France, Germany and, obviously, Spain representing the European Union (EU) developed countries; we also have added Poland as an emerging countries and new members of the (EU); representing Asia, we have considered Korea2; and finally we have included Mexico as a developing country that also covers the Latin-American community.

As a first step in our analysis, in Table I we offer a brief description of the cinema and theatre consumption habits in these seven countries. Following the Survey, these habits can be classified in four groups: never going to cinema or theatre, going 1 or 2 times a year, going 3 or 4 times a year and going more than 4 times a year

The first characteristic we can observe is that young people consumes much more cinema than theatre.

Second, Korea is notoriously different from the rest of countries. Young Koreans do not consume theatre or cinema. In the case of the last, 70.5 percent never go to the cinema and only a 7.4 percent go three or more times a year. The figures corresponding to theatre are more pessimistic: 88 percent never go and only 2 percent go three or more times.

Table I: Attendance to cinema and theatre

1 or 2 times a 3 or 4 times a More 4 times a Never year year year Cinema 5.4 19.4 22.6 52.6 France Theatre 66.3 27 4.4 2.3 Cinema 4 18.1 26.7 51.2 Germany Theatre 58.5 33.7 5.5 2.2 Cinema 70.5 21.9 3.7 4 Korea Theatre 88.1 10.2 1.2 0.5 Cinema 28.4 26.9 14.1 30.6 Mexico Theatre 60.3 30 6.6 3.1 Cinema 9.7 38 22.3 30 Poland Theatre 42.9 40.2 10 6.9 Cinema 7.4 16.1 18.5 57.9 Spain Theatre 54.4 34.9 7.7 3 Cinema 2.5 11.3 15.4 70.7 USA Theatre 47.2 38.8 8.4 5.6 Source: PISA 2000

2 We have rejected Japan because in our Survey there is not information about some relevant variables in our study. 4

Third, the different behaviour between cinema and theatre is higher in the case of the developed countries where pictures seem to be an important product of youth leisure.

Fourth, the consumption of cinema is specially important in USA: only a 2.5 percent of young Americans go to cinema while 70.7 percent do it four or more times. On the other hand, almost one half (47.2 percent) never goes to the theatre and a 14 percent go three or more times a year.

Fifth, we can identify a, more or less, homogeneous behaviour among developed European countries. The figures of non-attendants to cinema are quite small (from 4 percent in Germany to 7.4 percent in Spain) and more than a half of their young people go more than four times a year. In the case of theatre, between 55 percent (Spain) and 66 percent (France) never go and no more than 10 percent go more than three times a year.

Sixth, emerging European countries share some characteristics that make them different from other European countries and, in some sense, it could be identified as a soviet system inheritance. Thus, the figures of cinema attendance are smaller than the corresponding to developed countries (young people that go to the cinema more than four times a year represent between 30 and 40 percent), but we can identify a certain approximation path.

Finally, we discover serious differences in Mexico. Young people are a relevant market share in the case of cinema, although their attendance is lesser than in USA or Europe (almost a thirty percent of young Mexicans never go); however, these differences are narrower in the case of theatre, specially when we compare Mexico with the developed European countries where we can find very similar percentages.

In sum, we can identify the presence some relevant differences in the young people’s behaviour across the different countries considered.

THE MODEL

To analyse what factors determine these differences and, simultaneously, define the profiles of cinema and theatre attendants in each country we propose to estimate two multinomial ordered probit model, oriented to cinema and theatre consumption respectively, where the dependent variables are the attendance to cinema and theatre. Since these variables are made up of a scale of four ordered answers, multinomial probit (or logit) models do not 5 allow us to capture the ordinal nature of these variables. As Greene (2000) and Asworth and Johnson (1996) point up, multinomial ordered probit (or logit) models are the best methods of estimation in these cases.

In general terms, a multinomial ordered probit model could be presented as follows: y * = β X + ε where y* is not observable, X is a vector of independent variables and ε is normally distributed.

When you ask a person about his or her attendance to cinema (theatre), he or she would want to give you an answer according to his or her preferences (y*). However the PISA Survey only gives him or her four alternative responses. So, he or she has to select the closer answer to his or her preferences. Both in the cases of cinema and theatre, the structure of answers that we can observe in the survey (yi), and their relationship with the unobserved preferences (yi*) is:

yi = never = 0 if yi* ≤ 0

yi = 1 or 2 times a year = 1 if 0 < yi* ≤ µ1

yi = 3 or 4 times a year = 2 if µ1< yi* ≤ µ2

yi = More than 4 times a year = 3 if µ2< yi*

The µs are unknown parameters and they should be estimated together with the vector of parameters β.

The vector of explanatory variables (X) includes some socio-economic characteristics of the interviewee and her (his) family, with special attention to certain variables that can reflect the cultural environment of the family as a group. With all these variables we want to approximate both the individual’s process of taste formation and the family budget constraint. We can group these variables into the following categories: − Personal characteristics: includes gender, birthplace, city of residence, public (or private) school and performance in reading comprehension tests. − Familiar characteristics: includes parents’ work and level of studies. − Income variables: with this group of variables we try to approximate the income effect and we include some characteristics of the family equipment and cultural possessions. − Cultural environment of the family: in this case we try to find if there is a cultural joint consumption or attitude in the family. Of course, some other variables included in the previous group can also contribute to this information.

6 EMPIRICAL RESULTS

First of all, the correspondent likelihood ratios endorse the goodness of all of our estimations and all the µ parameters are statistically significant so we can confirm the existence of groups with different strength in the consumption of cinema a theatre in all the countries.

Due to the quantity and complexity of the results of these estimations, our strategy to comment them is to check if the principal hypothesis previously advanced are confirmed or not in general and then, we will analyse if there are significant behavioural differences among countries and between cinema and theatre inside each one of them (Tables II and III sum up the principal results).

• The educational variable, measured by the interviewee’s score in a reading comprehension test, seems to have a positive and significant influence, in the case of cinema, in almost all the countries. Only in the case of Korea the coefficient is not statistically significant. Hence, we can conclude that the best your educational performance the higher probability of attendance to cinema. In theatre the results are not so robust. Although one could expect that education has a more relevant role in the theatre taste formation process, we only observe a positive effect in the European countries. This variable is not significant in the American ones and, even more surprising, it has a negative effect in Korea. These results invite us to a further investigation trying to test if there are significant differences educational systems for example if European countries offer more theatre experiences at the school or, in the case of USA, if it is more linked to audiovisual culture.

• Other source of potential influence from the educational system is the presence of significant differences between interviewees belonging to public or private schools. We want to test if one of tem is more interested in cultural products than the other.We do not have obtained a general conclusion but, when the correspondent variable (PRIVSCHOOL) is statistically significant (in Spain, Germany and Poland for theatre, and Spain and Mexico for cinema) it has a positive sign. Then, we can conclude that, in this cases and specially in Spain, private schools are more interested in culture, due to their students have a higher probability of attendance to cinema or theatre.

• The variable OTHELANG, that covers the ability of speaking foreign languages at home, can also be linked to high educational level. It reinforces the effect of the

7 previous variables but only in few cases, as cinema in Spain or theatre in France and Poland.

• The familiar environment has an important influence on the young people’s consumption of cinema and theatre, as we have hypothesised above. Our set of variables that measure the frequency of talking about books, films or TV with the parents are very significant both in cinema and theatre and, in general terms, the have an increasing effect. Hence talking about cultural commodities seems to be a very good incentive to consume them, and this result is independent of the economic development or the peculiar background of the country.

• The educational level of the interviewee’s parents can be interpreted as proxies of cultural environment and heritage. In general terms, and when these variables are statistically significant, the father’s level of studies tends to have a positive influence on cinema and theatre consumption while the mother’s one tends to have a negative effect. Poland and Korea are rule exceptions and Mexico deserves a particular comment because both parents’ educational level have a positive, and increasing, effect in both goods. In sum, we can opine that, in the most developed western countries, fathers have a leading role in the taste formation process.

• Other variable related to cultural environment is CULPOSS, that represents the presence of some cultural possessions at home, and its effect is positive in all the countries.

• Adding the effect of the last three groups of variables we conclude that living a good cultural environment increases seriously the probability of cinema and theatre consumption.

• We also have tested the presence of a positive income effect using variables that define the presence of certain variables that pick up in the family equipment. In general terms these variables tend to have a positive impact reinforcing the presence of the income effect. This situation is remarkable in the case of having a personal room (ROOM) or being connected to the Internet (INTERNET) and in the case of cinema where they are always significant and have a positive effect on consumption. It is also remarkable that we cannot find any sign of income effect in the case of theatre consumption in USA and Germany and a very weaken sign in the case of France. Due to they are the richest countries of our sample it seems that, with the economic development, theatre losses its characteristic of an elite product 8

• We also interpret the presence of some positive and significant coefficients of the variable MWORK, linked to the mother’s wok, as a new symptom of a positive income effect.

• We have also tested the size of the city of residence influence on the probabilities of consuming both types of cultural products and both activities can be considered as urban behaviours. Moreover, if we look at the size of the coefficients, we conclude that young people who live in big cities (m ore than 1,000,000 inhabitants) have the highest probabilities of attending cinema and theatre, except in the case of theatre Spain where this probability belongs to young people living in small and middle towns (from 15,000 to 100,000 inhabitants). Korea is the unique exception because the probability of attendance to cinema and theatre is higher in rural areas.

Finally we must remark the presence of significant gender differences between good and countries. First, theatre is a women activity in all the analysed countries. Second, cinema presents all the possible results: in USA and Korea we do not identify any gender difference, while women have higher probability of attendance in Spain and Germany and men in France, Mexico and Poland.

APPENDIX 1 In this Appendix we define the variables used in this paper

A. Dependent Variables

CINEATTEND: Ordered discrete variable; it takes the following values: 1, if the interviewee never goes to the cinema 2, if the interviewee goes to the cinema 1 or 2 times a year 3, if the interviewee goes to the cinema 3 or 4 times a year 4, if the interviewee goes to the cinema more than 4 times a year

THETATTEND: Ordered discrete variable; it takes the following values: 1, if the interviewee never goes to the theatre 2, if the interviewee goes to the theatre 1 or 2 times a year 3, if the interviewee goes to the theatre 3 or 4 times a year 4, if the interviewee goes to the theatre more than 4 times a year

9 B. Independent Variables

MALE: Dummy variable; it takes value one when the interviewee is a man, and zero otherwise.

FPRIMARY: Dummy variable; it takes value one when the interviewee’s father has primary studies, and zero otherwise. FSECOND: Dummy variable; it takes value one when the interviewee’s father has secondary studies, FTERTIARY: Dummy variable; it takes value one when the interviewee’s father has tertiary studies, and zero otherwise.

MPRIMARY: Dummy variable; it takes value one when the interviewee’s mother has primary studies, and zero otherwise. MSECOND: Dummy variable; it takes value one when the interviewee’s mother has secondary studies, MTERTIARY: Dummy variable; it takes value one when the interviewee’s mother has tertiary studies, and zero otherwise.

FWORK: Dummy variable; it takes value one when the interviewee’s father works, and zero otherwise. MWORK: Dummy variable; it takes value one when the interviewee’s mother works, and zero otherwise.

ROOM: Dummy variable; it takes value one when the interviewee has not his own room, and zero otherwise.

BATHROOM1: Dummy variable; it takes value one when the interviewee has one bathroom at home, and zero otherwise. BATHROOM2: Dummy variable; it takes value one when the interviewee has two bathrooms at home, and zero otherwise. BATHROOM3: Dummy variable; it takes value one when the interviewee has three or more bathrooms at home, and zero otherwise.

COMPUTER: Dummy variable; it takes value one when the interviewee has one o more computers at home, and zero otherwise.

10 INTERNET: Dummy variable; it takes value one when the interviewee is not connected to the Internet at home, and zero otherwise.

CULTPOSS: This variable is PISA index related to “classical culture” in the family home. It measures the availability at home of classical literature, books of poetry and works of art .

PRIVSCHOOL: Dummy variable; it takes value one when the interviewee goes to a private school, and zero otherwise.

HABITAT1: Dummy variable; it takes value one when the interviewee lives in a small town (3,000 to about 15,000 people), and zero otherwise. HABITAT2: Dummy variable; it takes value one when the interviewee lives in a town (15,000 to about 100,000 people), and zero otherwise. HABITAT3: Dummy variable; it takes value one when the interviewee lives in a city (100,000 to about 1,000,000 people), and zero otherwise. HABITAT4: Dummy variable; it takes value one when the interviewee lives close to the centre of a city with over 1,000,000 people, and zero otherwise. HABITAT5: Dummy variable; it takes value one when the interviewee lives elsewhere in a city with over 1,000,000 people, and zero otherwise.

OTHLANG: Dummy variable; it takes value one when the family speaks at home other language most of the time, and zero otherwise.

TALKFAM1: Dummy variable; it takes value one when the interviewee discuss books, films or TV programmes with her (his) parents a few times a year. TALKFAM2: Dummy variable; it takes value one when the interviewee discuss books, films or TV programmes with her (his) parents about once a month. TALKFAM3: Dummy variable; it takes value one when the interviewee discuss books, films or TV programmes with her (his) parents several times a month. TALKFAM4: Dummy variable; it takes value one when the interviewee discuss books, films or TV programmes with her (his) parents several times a week.

READSCORE: Student performance in reading literacy. It has three dimensions: the type of reading task, the form and structure of the reading material and the use for which the text was constructed.

REFERENCES STIGLER, G. and BECKER, G. (1977), “De Gustibus Non Est Disputandum”, American Economic Review, 67, pp. 76-90. LÉVY-GARBOUA, L. and MONTMARQUETTE, C. (1996), “A Microeconometric Study of Theatre Demand”, Journal of Cultural Economics, 20, pp. 25-50. McCAIN, R. A. (1981), “Cultivation of Taste, Catastrophe Theory, and the Demand for Works of Art”, American Economic Review, 71, pp. 332-334. McCAIN, R. A. (1995), “Cultivation of Taste and Bounded Rationality: Some Computer Simulations”, Journal of Cultural Economics, 19, pp. 1-15. McCAIN, R. A. (2003), “Taste Formation” in TOWSE, T. (ed.), A Handbook of Cultural Economics, Edward Elgar, Cheltenham, pp. 445-450. LÉVY-GARBOUA, L. (2003), “Demand” in TOWSE, T. (ed.), A Handbook of Cultural Economics, Edward Elgar, Cheltenham, pp. 201-213. MORRISON, W. G. and WEST, E. G. (1986), “Child Exposure to the Performing Arts: The Implications for Adult Demand”, Journal of Cultural Economics, 10, pp. 17-24. ABBÉ-DECARROUX, F. (1995), “Demande Artistique et préférences endogenes”, Revue Economique, 46, pp.983-992. OECD (2001), Knowledge and Skills for Life. First Results from PISA 2000, OECD Publications, Paris ASHWORTH, J. and JOHNSON, P. (1996), “Sources of ‘Value for Money’ for Museum Visitors: Some Survey Evidence”, Journal of Cultural, 20, pp 67-83. BAUMOL, W. and BOWEN, W. (1966), Performing Arts - The Economic Dilemma, The Twentieth Century Found, New York. O’HAGAN, J. (1996), “Access to and Participation in the Arts: The Case of Those with Low Incomes/Educational Attainment”, Journal of Cultural Economics, 20, pp.269-282. TOWSE, R. (1994), “Achieving Public Policy Objectives in the Arts and Heritage” en PEACOCK, A., and RIZZO, I. eds. (1994), Cultural Economics and Cultural Policies, Kluwer Academic Publishers, Dordrecht. THROSBY, D. and WITHERS, G. (1979), The Economics of Performing Arts, Edward Arnold, London.

12 TABLE II. ATTENDANCE TO CINEMA

SPAIN FRANCE GERMANY MEXICO POLAND USA KOREA Coef. t Coef. t Coef. t Coef. t Coef. t Coef. t Coef. t MALE -0,106 -3,29 0,089 2,35 -0,079 -2,13 0,090 2,40 0,091 2,19 -0,068 -1,36 -0,011 -0,29 FPRIMARY 0,054 0,91 -0,042 -0,41 0,375 2,47 0,043 0,72 -0,039 -0,13 0,052 0,31 -0,083 -0,63 FSECOND 0,131 2,09 -0,057 -0,82 0,178 2,52 0,234 3,61 -0,134 -1,94 0,080 0,98 -0,079 -0,67 FTERTIARY 0,154 2,17 0,020 0,25 0,154 2,02 0,333 4,24 0,022 0,23 0,203 2,15 0,127 1,08 MPRIMARY -0,111 -1,73 -0,205 -1,83 -0,194 -1,42 0,267 4,49 0,127 0,30 -0,159 -0,87 -0,064 -0,40 MSECOND -0,042 -0,62 -0,010 -0,12 -0,119 -1,56 0,537 7,98 0,088 1,10 -0,139 -1,31 0,182 1,18 MTERTIARY -0,075 -0,93 0,009 0,10 -0,015 -0,17 0,506 5,66 0,184 1,84 -0,221 -1,92 0,298 1,84 FWORK 0,009 0,17 0,064 1,18 0,058 1,01 0,043 0,83 0,031 0,71 0,011 0,17 0,058 1,10 MWORK 0,052 1,62 0,175 4,33 0,152 3,82 0,024 0,60 0,155 3,68 0,155 2,81 -0,035 -0,90 OTHELANG 0,042 0,28 0,196 1,95 -0,028 -0,32 0,660 1,47 1,660 3,73 -0,044 -0,51 * * ROOM -0,086 -2,34 -0,231 -4,41 -0,179 -2,71 -0,196 -5,16 -0,221 -4,51 -0,167 -2,42 -0,188 -2,75 INTERNET -0,188 -4,41 -0,162 -3,38 -0,153 -3,92 -0,353 -4,75 -0,265 -4,42 -0,210 -3,16 -0,136 -3,04 COMPUTER 0,153 4,18 0,066 1,53 0,199 3,49 0,301 5,24 0,241 5,21 -0,094 -1,22 0,092 1,45 BATHROOM1 0,202 2,13 -0,050 -0,60 -0,018 -0,10 0,111 0,88 0,098 1,17 0,125 0,40 0,072 0,61 BATHROOM2 0,214 2,24 0,004 0,05 0,100 0,55 0,247 1,92 0,210 2,28 0,177 0,57 0,116 0,96 BATHROOM3 0,259 2,43 0,181 1,33 0,131 0,71 0,472 3,21 0,530 3,96 0,295 0,94 -0,035 -0,17 HABITAT1 0,232 2,06 0,354 4,52 -0,033 -0,43 0,000 0,00 -0,219 -1,76 0,101 1,01 -0,531 -4,15 HABITAT2 0,454 4,08 0,377 5,16 0,047 0,64 0,157 2,30 -0,141 -1,18 0,176 1,78 -0,448 -3,71 HABITAT3 0,475 4,27 0,523 6,23 0,083 1,01 0,541 7,62 0,049 0,41 0,316 3,05 -0,414 -3,79 HABITAT4 0,266 1,91 0,617 2,29 0,218 1,48 0,690 6,96 0,418 2,98 0,618 3,92 -0,476 -4,11 HABITAT5 0,597 4,61 0,479 3,74 0,339 2,85 0,496 5,46 0,418 2,15 0,424 3,20 -0,422 -3,79 PRIVSCHOOL 0,079 2,23 -0,035 -0,77 0,142 1,62 0,247 3,71 -0,035 -0,26 0,089 0,78 0,032 0,81 CULTPOSS 0,094 5,26 0,064 3,09 0,037 1,85 0,124 5,88 0,128 5,32 0,089 3,41 0,185 7,71 TALKFAM1 0,155 2,20 0,121 1,52 0,165 2,88 0,199 2,99 0,169 2,39 0,149 1,74 0,554 10,14 TALKFAM2 0,329 4,57 0,305 4,06 0,246 4,09 0,313 4,25 0,192 2,60 0,328 3,42 0,534 8,16 TALKFAM3 0,380 5,81 0,365 5,36 0,338 5,86 0,481 6,83 0,318 4,51 0,427 4,88 0,561 9,73 TALKFAM4 0,570 8,83 0,462 6,77 0,478 7,35 0,491 7,10 0,436 6,31 0,417 5,26 0,618 10,83 READSCORE 0,003 11,15 0,001 5,41 0,001 4,60 0,002 6,43 0,001 6,05 0,001 2,24 -0,001 -1,59 cut_1 0,645 -0,321 -0,769 0,952 -0,42 -1,279 0,572 cut_2 1,466 0,687 0,292 1,921 1 -0,272 1,549 cut_3 2,047 1,351 1,075 2,437 1,668 0,331 1,893 N 5836 3955 4233 4068 3200 2814 4813 LRchi2 929,88 425,26 371,2 2086,86 698,23 234,74 549,03

13 TABLE III. ATTENDANCE TO THEATRE

SPAIN FRANCE GERMANY MEXICO POLAND USA KOREA Coef. z Coef. z Coef. z Coef. z Coef. z Coef. z Coef. z MALE -0,218 -6,820 -0,138 -3,330 -0,168 -4,320 -0,077 -1,960 -0,093 -2,190 -0,236 -5,250 -0,134 -2,640 FPRIMARY 0,070 1,120 0,164 1,360 0,090 0,540 0,081 1,170 -0,182 -0,520 0,075 0,440 -0,196 -1,060 FSECOND 0,067 1,030 0,160 1,940 0,073 0,940 0,214 2,920 -0,017 -0,230 -0,033 -0,410 -0,050 -0,320 FTERTIARY 0,085 1,200 0,313 3,510 0,211 2,590 0,243 2,880 0,125 1,320 0,132 1,470 0,220 1,410 MPRIMARY -0,120 -1,770 -0,215 -1,640 0,173 1,150 0,121 1,760 -0,846 -1,420 -0,548 -2,740 0,304 1,050 MSECOND -0,063 -0,890 -0,244 -2,570 -0,009 -0,110 0,146 1,910 0,050 0,600 -0,105 -1,040 0,495 1,750 MTERTIARY 0,026 0,330 -0,160 -1,560 0,165 1,810 0,091 0,970 0,181 1,800 0,050 0,460 0,701 2,430 FWORK -0,028 -0,560 -0,005 -0,080 -0,095 -1,540 -0,075 -1,330 -0,009 -0,190 0,033 0,550 0,202 2,700 MWORK 0,053 1,670 -0,014 -0,310 0,038 0,900 0,093 2,270 0,103 2,360 -0,038 -0,750 -0,036 -0,700 OTHELANG 0,352 2,450 0,061 0,540 -0,006 -0,060 -0,237 -0,570 -0,075 -0,260 -0,028 -0,350 * * ROOM -0,031 -0,820 -0,018 -0,300 0,077 1,070 -0,218 -5,400 -0,186 -3,590 -0,092 -1,370 -0,231 -2,320 INTERNET -0,134 -3,390 -0,147 -2,970 0,006 0,140 -0,297 -4,420 0,014 0,230 -0,093 -1,470 -0,044 -0,740 COMPUTER 0,105 2,800 -0,034 -0,710 0,034 0,540 0,154 2,660 0,139 2,930 0,038 0,510 0,136 1,520 BATHROOM1 0,018 0,180 0,098 0,990 -0,001 0,000 0,210 1,380 -0,123 -1,410 -0,297 -0,920 0,321 1,760 BATHROOM2 0,014 0,140 0,130 1,250 0,016 0,080 0,287 1,850 -0,054 -0,560 -0,242 -0,750 0,413 2,240 BATHROOM3 -0,027 -0,250 0,256 1,830 -0,115 -0,550 0,123 0,740 0,070 0,530 -0,234 -0,720 0,436 1,660 HABITAT1 0,619 4,810 -0,088 -0,950 -0,159 -1,910 -0,122 -1,570 -0,374 -2,910 0,171 1,740 -0,371 -2,200 HABITAT2 0,564 4,430 -0,004 -0,050 -0,121 -1,490 -0,038 -0,500 -0,272 -2,230 0,136 1,410 -0,323 -2,040 HABITAT3 0,363 2,860 -0,067 -0,700 0,069 0,790 0,216 2,760 -0,017 -0,140 0,213 2,140 -0,293 -2,070 HABITAT4 0,322 2,160 0,641 2,850 0,295 2,070 0,100 1,000 0,311 2,220 0,226 1,640 -0,301 -2,020 HABITAT5 0,244 1,720 0,109 0,820 0,093 0,760 0,609 6,330 -0,110 -0,580 0,360 2,940 -0,246 -1,720 PRIVSCHOOL 0,128 3,760 0,041 0,820 0,232 2,810 0,022 0,350 0,400 3,130 -0,005 -0,050 0,036 0,690 CULTPOSS 0,137 7,450 0,229 9,900 0,244 11,240 0,215 9,700 0,159 6,320 0,286 11,830 0,278 8,200 TALKFAM1 0,029 0,380 -0,096 -0,980 0,138 2,110 0,172 2,230 0,149 1,940 0,389 4,260 0,510 6,860 TALKFAM2 0,039 0,490 0,067 0,750 0,118 1,750 0,276 3,310 0,359 4,560 0,530 5,500 0,512 5,880 TALKFAM3 0,073 1,030 0,145 1,790 0,238 3,720 0,349 4,410 0,336 4,470 0,484 5,450 0,575 7,490 TALKFAM4 0,153 2,190 0,133 1,650 0,294 4,230 0,336 4,290 0,517 7,070 0,545 6,580 0,542 7,090 READSCORE 0,002 6,710 0,002 5,770 0,001 6,400 0,000 0,380 0,001 3,100 0,000 0,830 -0,002 -4,160 cut_1 1,374 1,155 1,117 0,626 0,355 0,183 1,537 cut_2 2,572 2,366 2,440 1,803 1,625 1,481 2,594 cut_3 3,247 2,921 3,087 2,455 2,214 2,036 3,125 n 5836 3955 4233 4068 3200 2814 4813 LRchi2 481,780 460,070 560,130 773,110 512,140 489,160 430,110

14 TABLE IV. DESCRIPTIVE STATISTICS

SPAIN FRANCE GERMANY MEXICO POLAND EEUU KOREA Variable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. MALE 0,489 0,500 0,491 0,500 0,485 0,500 0,500 0,500 0,512 0,500 0,471 0,499 0,554 0,497 FPRIMARY 0,333 0,471 0,051 0,220 0,018 0,133 0,316 0,465 0,004 0,064 0,025 0,157 0,113 0,317 FSECOND 0,345 0,475 0,507 0,500 0,530 0,499 0,334 0,472 0,731 0,443 0,527 0,499 0,620 0,485 FTERTIARY 0,223 0,416 0,326 0,469 0,327 0,469 0,196 0,397 0,133 0,339 0,311 0,463 0,237 0,425 MPRIMARY 0,395 0,489 0,047 0,211 0,023 0,151 0,381 0,486 0,002 0,047 0,024 0,152 0,152 0,359 MSECOND 0,364 0,481 0,571 0,495 0,674 0,469 0,339 0,474 0,754 0,431 0,601 0,490 0,705 0,456 MTERTIARY 0,163 0,369 0,304 0,460 0,201 0,401 0,127 0,333 0,151 0,358 0,307 0,461 0,120 0,325 FWORK 0,885 0,319 0,854 0,353 0,886 0,318 0,843 0,363 0,667 0,472 0,782 0,413 0,808 0,394 MWORK 0,541 0,498 0,665 0,472 0,716 0,451 0,388 0,487 0,595 0,491 0,727 0,445 0,375 0,484 OTHELANG 0,011 0,105 0,039 0,193 0,050 0,219 0,002 0,047 0,005 0,071 0,104 0,305 * * ROOM 0,235 0,424 0,153 0,360 0,079 0,270 0,526 0,499 0,218 0,413 0,146 0,354 0,110 0,313 INTERNET 0,761 0,426 0,736 0,441 0,587 0,492 0,873 0,333 0,823 0,382 0,327 0,469 0,396 0,489 COMPUTER 0,672 0,469 0,653 0,476 0,883 0,322 0,245 0,430 0,419 0,494 0,810 0,393 0,847 0,360 BATHROOM1 0,413 0,492 0,651 0,477 0,431 0,495 0,587 0,492 0,688 0,464 0,256 0,436 0,668 0,471 BATHROOM2 0,449 0,497 0,261 0,439 0,398 0,490 0,299 0,458 0,210 0,407 0,455 0,498 0,289 0,453 BATHROOM3 0,111 0,315 0,036 0,185 0,162 0,369 0,092 0,289 0,040 0,196 0,284 0,451 0,014 0,118 HABITAT1 0,211 0,408 0,215 0,411 0,274 0,446 0,210 0,407 0,150 0,357 0,267 0,443 0,074 0,262 HABITAT2 0,321 0,467 0,529 0,499 0,433 0,496 0,244 0,430 0,397 0,489 0,327 0,469 0,099 0,298 HABITAT3 0,358 0,479 0,147 0,354 0,171 0,377 0,286 0,452 0,339 0,473 0,226 0,419 0,349 0,477 HABITAT4 0,036 0,185 0,007 0,084 0,021 0,143 0,071 0,257 0,068 0,251 0,044 0,205 0,150 0,357 HABITAT5 0,055 0,228 0,034 0,180 0,036 0,185 0,068 0,252 0,018 0,131 0,066 0,248 0,297 0,457 PRIVSCHOOL 0,401 0,490 0,217 0,412 0,050 0,217 0,155 0,362 0,026 0,160 0,058 0,233 0,507 0,500 CULTPOSS 0,182 0,955 -0,304 0,996 0,034 0,984 -0,550 1,003 0,149 0,899 -0,165 1,040 0,208 0,904 TALKFAM1 0,121 0,326 0,106 0,308 0,220 0,415 0,255 0,436 0,195 0,396 0,163 0,369 0,193 0,395 TALKFAM2 0,120 0,325 0,149 0,356 0,194 0,396 0,151 0,358 0,166 0,372 0,122 0,327 0,107 0,309 TALKFAM3 0,290 0,454 0,312 0,463 0,266 0,442 0,220 0,414 0,233 0,423 0,207 0,405 0,164 0,370 TALKFAM4 0,403 0,490 0,341 0,474 0,165 0,371 0,258 0,438 0,281 0,449 0,387 0,487 0,170 0,375 READSCORE 496,306 80,233 505,232 87,114 506,796 95,182 432,945 81,488 476,425 91,430 501,885 98,301 522,454 65,739